[all] Collecting Virtual Midcycle Best Practices

Mark Goddard mark at stackhpc.com
Tue Mar 10 09:45:22 UTC 2020


On Tue, 10 Mar 2020 at 09:17, Thierry Carrez <thierry at openstack.org> wrote:
>
> Kendall Nelson wrote:
> > I wanted to collect best practices and pitfalls to avoid wrt projects
> > experiences with virtual midcycles. I know of a few projects that have
> > done them in the past and with how travel is hard for a lot of people
> > right now, I expect more projects to have midcycles. I think it would be
> > helpful to have all of the data we can collect in one place for those
> > not just new to virtual midcycles but the whole community.
> > [...]
>
> Also interested in feedback from teams that had virtual PTGs in the past
> (keeping all possibilities on the table). I think Kolla, Telemetry and a
> few others did that.

Kolla has now had two virtual PTGs. Overall I think they went fairly
well, particularly the most recent one. We tried Zoom, then moved to
Google Meet. I forget the problems with Zoom. There were inevitably a
few teething problems with the video, but I think we worked it out
after 15-20 minutes. Etherpad for Ussuri vPTG here:
https://etherpad.openstack.org/p/kolla-ussuri-ptg.

Without seeing people's faces it can be hard to ensure everyone keeps
focussed. It's quite rare for the whole room to be focussed at
physical discussions though.

Going around the room giving short intros helps to get people talking,
and it may be better to do these ~1 hour in as people may miss the
start. Directing questions at non-cores can help overcome that pesky
imposter syndrome. Keeping video on definitely helps with engagement,
up to the point where it impacts audio quality.

There was also the Denver PTG where the PTL and a number of cores were
remote where we struggled to make any progress. I think there were a
few reasons for this. The fixed time of the PTG was not optimal for
many remote attendees living in Europe or Asia. When there are a
number of participants in one location, it can be easy to forget to
direct speech at the microphone, allow time for remote callers to ask
questions/respond etc. This makes it difficult and frustrating for
them to join in, making it easier to get distracted and drop off.

Not too much hard data in there, but hopefully a feel for how it went for us.

>
> --
> Thierry Carrez (ttx)
>



More information about the openstack-discuss mailing list